US6018523A - Switching networks having improved layouts - Google Patents
Switching networks having improved layouts Download PDFInfo
- Publication number
- US6018523A US6018523A US08/955,685 US95568597A US6018523A US 6018523 A US6018523 A US 6018523A US 95568597 A US95568597 A US 95568597A US 6018523 A US6018523 A US 6018523A
- Authority
- US
- United States
- Prior art keywords
- network
- switching
- column
- switching circuits
- grid
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Lifetime
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L49/00—Packet switching elements
- H04L49/10—Packet switching elements characterised by the switching fabric construction
- H04L49/104—Asynchronous transfer mode [ATM] switching fabrics
- H04L49/105—ATM switching elements
- H04L49/106—ATM switching elements using space switching, e.g. crossbar or matrix
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q11/00—Selecting arrangements for multiplex systems
- H04Q11/04—Selecting arrangements for multiplex systems for time-division multiplexing
- H04Q11/0428—Integrated services digital network, i.e. systems for transmission of different types of digitised signals, e.g. speech, data, telecentral, television signals
- H04Q11/0478—Provisions for broadband connections
Definitions
- the present invention relates to switching networks. More particularly, the invention relates to a reduced-area layout for switching networks.
- Switching networks are useful as fast circuits for routing data.
- Switching networks consist of input terminals, a plurality of switching circuits, interconnections or links, and output terminals.
- a switching circuit is operable to route two signals appearing at its inputs to either of two outputs. The route through the switching circuit is dictated by an external controller.
- Switching networks are useable, for example, in telephone switchboards, the switching fabric of Asynchronous Transfer Mode (ATM) switches and the like.
- ATM Asynchronous Transfer Mode
- a given network has, however, a characteristic number of switching circuits as a function of the number of network inputs. But, there are at least two degrees of freedom available for use in developing a network layout.
- a first is the layout of network switching circuits, i.e., their placement on a grid relative to one another.
- a second is the layout of the interconnects linking such switching circuits. In fact, those two aspects of a network layout may be of paramount importance in determining network area requirements.
- the present invention is directed to a layout for a switching network having N ⁇ log N switching circuits.
- That switching network which is attributed to V. Benes and A. Waksman, is commonly referred to as the "Benes" switching network. See, V. E. Benes, Mathematical Theory of Connecting Networks and Telephone Traffic, (Academic Press, 1965); A. Waksman, "A Permutation Network,” J. ACM, v. 15, no. 1, pp. 159-63 (January 1968).
- Those publications, and any other publications or patent applications referenced in this specification, are incorporated by reference herein.
- the area required for implementing a network can be determined by developing a grid layout for the network.
- the grid layout encompasses the various interconnect permutations required within the network.
- the Benes switching network can be laid-out in an area of 3N 2 or fewer grid units for a network of N inputs.
- the switching circuits comprising the network are arranged in columns, but not necessarily in rows, as in conventional arrangements.
- the present arrangement advantageously facilitates using links, for interconnecting switching circuits, which are "slanted,” in substantial portion, relative to the arrangement of switching circuits in the columns.
- the links are slanted by 45°. Conventional layouts of the Benes network do not use slanted links.
- the links used in conjunction with the present invention are either straight slanted lines, or have at most one bend, in comparison with prior art layouts having links with as many as four bends. Such bends increase network layout area requirements and complicate fabrication.
- a layout of the Benes switching network in accordance with the present invention advantageously requires significantly less grid area than conventional layouts of same.
- FIG. 1 shows a simplified illustration of a switching circuit C
- FIG. 2 shows a portion of an exemplary grid for laying out a network and a representation of grid area
- FIG. 3 shows a conceptual illustration of the Benes switching network
- FIG. 4 shows the conceptual structural representation of a layout of one half of the Benes network H n+1 , in accordance with the present invention
- FIGS. 5a-5d show a layout of H 1 -H 4 in accordance with the present invention.
- FIG. 7 shows a simplified diagram of a portion of an asynchronous transfer mode network
- FIG. 8 shows a simplified diagram of a switch.
- An exemplary switching network has N input terminals and N output terminals and a plurality of switching circuits and interconnects or links. If N signals, representing N values, such as, for example, real numbers or address values, are fed into the input terminals, the same signals appear on the output terminals.
- Switching networks can be implemented to process electronic, optical or other types of signals. For clarity and ease of presentation, such signals will hereinafter be referred to as "numbers.” Further, as used herein, the term "vector" is defined as a sequence of signals. It will be appreciated that the physical implementation of the network will vary depending upon the signal-type being processed, such as, for example, an electrical or an optical signal type.
- FIG. 1 shows a simplified illustration of a switching circuit C.
- a switching circuit is well known in the art, and contains logic circuits, memory (for synchronization), and the like.
- the switching circuit C has two input terminals 6, 8 and two output terminals 10, 12, as illustrated in FIG. 1. If two numbers x, y are fed to the two inputs 6, 8, the same numbers x, y emerge on the output terminals 10, 12.
- a controller not shown, determines the routing through the switching, dictating on which output 10, 12 a given input number will appear.
- designations for numbers (signals) delivered to the terminals, and the terminals receiving such numbers will be used interchangeably.
- the networks described herein also have fixed interconnects or links, such as the fixed interconnects 4a-d shown in FIG. 1.
- Such interconnects are normally implemented, for example, as conductive traces patterned on and throughout integrated circuits, or as optical waveguides in optical circuits. All such interconnects are directed links connecting a network input terminal, not shown, or a switching circuit output terminal to a switching circuit input terminal, or a network output terminal, not shown.
- a grid can be used to assist in laying out a network.
- a grid consists of a first and a second set of straight lines. In each of such sets, the lines are parallel. It is assumed herein that the distance between any two adjacent parallel lines is the same as the distance between any other two adjacent parallel lines, though such distance can vary. That distance is assumed to be equal to one unit, referred to herein as a "grid segment.”
- the lines of the first set may lie in any non-zero angle with respect to the lines of the second set. For clarity of presentation, that angle is assumed herein to be ninety degrees, i.e., the lines of the first set are orthogonal to those of the second set.
- the grid is a square.
- the grid points are the points of intersections of the lines of the first set with the lines of the second set.
- FIG. 2 shows a portion of such a grid.
- the first and second sets of straight lines are oriented vertically as “grid columns,” and horizontally as “grid rows.”
- the grid portion shown is defined by grid columns co19-col14 and grid rows row11-row14.
- the grid area encompassed by a region bounded by grid columns col9-col14 (side a), and grid rows row11-row 14 (side b) is fifteen grid units (GU).
- a grid unit GU is not arbitrarily small.
- a grid unit has a minimum size dictated by the state-of-the-art in device fabrication, particularly in terms of minimum feature size, prevailing at any given point in time. For example, assume it is determined that a sorting network requires an area of G grid units. It will be appreciated that the absolute area, as measured in square microns or square millimeters, required to implement such a network using technology current in, say 1970, is significantly greater than the area required for implementation using technology current in 1997.
- the layout areas presented herein are described in terms of "grid units.” It should now be appreciated that a grid unit is a relative measure, and not an absolute measure, of layout area.
- the area required for a layout of a switching network is the area (measured in grid units) of the smallest rectangle that contains such a layout.
- the grid area of a layout is defined as the least number of grid units in a rectangle that encompasses the layout.
- rectangle R contains at least [(a-1) ⁇ (b-1)]-1 grid points and at most (a+1) ⁇ (b+1) grid points, no matter how the rectangle R is oriented with respect to the grid.
- the sides of R need not be parallel to the grid lines and the quantities a and b need not be integers. When a and b are large, the number of grid points is approximately equal to the area of the rectangle in grid units.
- a switching network by a directed graph, wherein switching circuits and network terminals are considered to be vertices, and the interconnects or links are directed edges.
- the following rules and definitions will be used herein with regard to embedding such a graph in a grid.
- the vertices of a graph are mapped to grid-points, with no more than one vertex per grid-point.
- every edge of the directed graph is represented by a path in the grid, but no grid edge is used in more than one such path.
- Such a mapping of the graph edges to grid paths is referred to herein as "edge-disjoint" mapping.
- Two paths may share an intermediate grid-point, but they must cross at that point, i.e., no "knock-knee” or change in direction is allowed.
- the Benes switching network utilizes N ⁇ log N switching circuits, and has a network depth of 2 log N.
- the Benes switching network S n consists of two Butterfly networks, H n 1 , H n 2 , well known to those skilled in the art, that are connected "back-to-back.”
- One known layout of the Butterfly network is the layout of D. S. Wise. See D. S. Wise, "Compact Layouts of Banyan/FFT Networks," VLSI Sytems and Computations, (G. Steele et al., eds.), pp. 186-95, (Computer Science Press, 1981).
- Wise's layout of the Butterfly network is not directly applicable for use in laying out the Benes switching network; i.e., one cannot simply use Wise's layout for connecting two butterfly networks, back-to-back, to obtain a Benes network.
- Wise layout is for the case wherein the network inputs and outputs are vertices (of degree 2), and the routing paths are vertex disjoint.
- Wise's layout includes knock-knees.
- the inputs are edges (or vertices of degree 1), and the routing paths are only required to be edge-disjoint.
- a layout of the Butterfly network has been discovered that is suitable for back-to-back connection and satisfies the other previouslydescribed rules and definitions. Such a layout can therefore be used to construct a layout of the Benes network.
- H n has N input edges and N/2 output vertices.
- Such a layout is, in fact, the Butterfly network for (n-1) with two input edges attached to each of the input vertices. Reflecting H n about its output vertices to effectively create a back-to-back connection, yields a layout of the Benes network S n with N input/output edges.
- FIG. 4 A conceptual structural representation of H n+1 , in accordance with the present invention, is shown in FIG. 4. Description of the recursive structure of the "half" net H n is facilitated by examining network construction and related elements in terms of the "n+1st" level.
- H n+1 comprises two copies of the net H n , H n a and H n b , placed one above the other, a column OSC n+1 of switching circuits, and channel linking CL for bringing together an output from H n a and an output from H n b at each of the switching circuits in column OSC n+1 .
- FIG. 5b shows the layout of H 2 .
- H 2 includes two copies of H 1 , i.e., H 1 a , H 1 b , disposed one above the other, and a final column OSC n of 2 switching circuits.
- the outputs from H n a and H 1 b are routed, via channel linking, to the switching circuits in column OSC n .
- the layout is described recursively, and applies generally to FIGS. 5b-5d.
- H n+1 comprises two copies of H n , i.e., H n a and H n b , situated one on top of the other.
- H 4 comprises two copies of H 3 .
- the gap g has a dimension of ⁇ 2 ⁇ (2 n-1 +1) grid segments.
- the final column OSC n+1 of switching circuits of H n+1 is positioned so that the 2 n switching circuits contained in OSC n+1 are located on the same rows as the middle 2 n network inputs.
- the switching circuits in OSC 4 are disposed, respectively, in the same rows as network inputs X 5 -X 12 .
- the spacing between adjacent switching circuits in OSC n+1 and adjacent network inputs is ⁇ 2 grid segments.
- the column OSC n+1 of switching circuits is disposed ⁇ 2 ⁇ 3 ⁇ (2 n-2 ) grid segments to the right of the columns OSC n in H n a and H n b . Note that unlike Wise's arrangement of the Butterfly network, in the present layout, the switching circuits from the various columns do not align into rows.
- each switching circuit in the column OSC n of H n a is connected by a straight slanted channel link CL to the upper input of each of the lower half of the switching circuits in column OSC n+1 .
- the upper output of each switching circuit in the column OSC n of H n b is connected by a straight slanted channel link CL to the lower input of each of the upper half of the switching circuits in column OSC n+1 .
- the lower output d of switching circuit SW17 is connected to upper input a of SW29
- the upper output c of switching circuit SW21 is connected to lower input b of SW25.
- the upper output of each switching circuit in the column OSC n of H n a is connected to the upper input of each of the upper half of the switching circuits in column OSC n+1 in reverse order.
- the lower output of each switching circuit in the column OSC n of H n b is connected to the lower input of each of the lower half of the switching circuits in column OSC n+1 in reverse order.
- the upper output c of SW17 is connected to the upper input of SW28.
- the lower output d of SW21 is connected to the lower input b of SW32.
- the channel links from H n a bends, preferably, at a row located ( ⁇ 2)/2 grid segments above the upper-most switching circuit in column OSC n of H n a .
- the bends occur, preferably, at a row located ( ⁇ 2)/2 grid segments below the lowermost switching circuit in column OSC n of H n b .
- the network S n is symmetric about an axis 2--2 passing vertically through the center of column OSC n of switching circuits.
- To the left of the axis 2--2 is a first copy of the butterfly network, H n ; to the right is a second copy of it.
- the column OSC n of 2 n-1 switching circuits is common to both copies of H n .
- each butterfly H n does not separately contribute its column OSC n of 2 n-1 switching circuits to the Benes network.
- the height of the half network H n (or S n ) is ⁇ 2 ⁇ (2 n -1) grid segments.
- the width increases by ⁇ 2 ⁇ 3 ⁇ (2 n-3 ) grid segments to obtain H n from H n-1 .
- the width of H n is ⁇ 2 ⁇ (3 ⁇ (2 n-2 )-1) grid segments.
- the grid area for a layout of the Benes network for N inputs in accordance with the present invention, has an upper bound grid area of 3N 2 grid units.
- an "upper bound" area of 3N 2 grid units means that the switching network can be implemented in 3N 2 or fewer grid units. Thus, in the present context, it is advantageous to reduce the upper bound. It should be appreciated, however, that in other less preferred embodiments of the present invention, the Benes switching network can be implemented in an area greater than 3N 2 grid units.
- the location and spacing of the columns, bends, and the like, described above allows the links within the network to be slanted at 45 degrees relative to a column defined by the network inputs (or relative to the various columns into which the switching circuits are arranged).
- the aforementioned arrangement obtains the minimum layout area according to the present invention.
- the links may be nonorthogonal to such columns, but not at 45°, or the links may contain more than one bend.
- the location of the bend in the links, or the position of the various columns of switching circuits may vary from the preferred embodiments described above. Layouts obtained using such other embodiments will have grid areas greater than 3N 2 grid units. It is believed, however, that many of such embodiments will still represent an improvement, in terms of grid area requirements, over prior art layouts of the Benes network. Such embodiments are contemplated to be within the scope of the present invention.
- Switching networks such as the Benes switching network
- the improved layouts described herein may be used advantageously, in the same applications, to minimize the area required to physically implement the switching network on one or more integrated circuits.
- FIG. 7 shows a portion of an ATM network having nodes N1-N5.
- a single source S1 delivers data onto the network at N1 for delivery to other nodes in the network, such as nodes N2-N5.
- Each of the nodes has an ATM node switch, not shown in FIG. 7, for routing data to other nodes in the network.
- FIG. 8 A simplified diagram of an ATM node switch according to the present invention is shown in FIG. 8.
- data packets D1-DN are received by the ATM node switch and stored in buffers for preprocessing in preprocessing stage PS.
- the data packets include routing or address data indicating the intended destination of the information content.
- decisions are made by known methods concerning the order of data flow into switch fabric SF, which is a switching network SN having an improved layout as described herein. From the buffers, the data is delivered synchronously into the switch fabric SF, and routed to the appropriate node.
Abstract
Description
Claims (12)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US08/955,685 US6018523A (en) | 1997-10-22 | 1997-10-22 | Switching networks having improved layouts |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US08/955,685 US6018523A (en) | 1997-10-22 | 1997-10-22 | Switching networks having improved layouts |
Publications (1)
Publication Number | Publication Date |
---|---|
US6018523A true US6018523A (en) | 2000-01-25 |
Family
ID=25497200
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US08/955,685 Expired - Lifetime US6018523A (en) | 1997-10-22 | 1997-10-22 | Switching networks having improved layouts |
Country Status (1)
Country | Link |
---|---|
US (1) | US6018523A (en) |
Cited By (23)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6185220B1 (en) * | 1998-06-15 | 2001-02-06 | Lucent Technologies, Inc. | Grid layouts of switching and sorting networks |
US6721311B1 (en) | 1999-08-10 | 2004-04-13 | Colorado Seminary | Self-routing permutation networks based on de Bruijn digraphs |
US20040155725A1 (en) * | 2003-02-06 | 2004-08-12 | Com Dev Ltd. | Bi-planar microwave switches and switch matrices |
US7113506B1 (en) * | 1999-06-08 | 2006-09-26 | Cisco Technology, Inc. | Butterfly network with switches set for two node disjoint paths and method for forming the paths |
US20070025257A1 (en) * | 2003-09-30 | 2007-02-01 | Fang Xu | Efficient switching architecture with reduced stub lengths |
US20090100314A1 (en) * | 2007-10-15 | 2009-04-16 | Coreoptics Inc. | Modification of error statistics behind equalizer to improve inter-working with different fec codes |
US20100180048A1 (en) * | 2009-01-09 | 2010-07-15 | Microsoft Corporation | Server-Centric High Performance Network Architecture for Modular Data Centers |
US20110080855A1 (en) * | 2009-10-01 | 2011-04-07 | Hei Tao Fung | Method for Building Scalable Ethernet Switch Network and Huge Ethernet Switch |
US7982497B1 (en) | 2010-06-21 | 2011-07-19 | Xilinx, Inc. | Multiplexer-based interconnection network |
US20110202682A1 (en) * | 2010-02-12 | 2011-08-18 | Microsoft Corporation | Network structure for data center unit interconnection |
US8098081B1 (en) | 2010-06-21 | 2012-01-17 | Xilinx, Inc. | Optimization of interconnection networks |
US20130124918A1 (en) * | 2002-12-02 | 2013-05-16 | Marvell World Trade Ltd. | Self-reparable semiconductor and method thereof |
US8665727B1 (en) | 2010-06-21 | 2014-03-04 | Xilinx, Inc. | Placement and routing for a multiplexer-based interconnection network |
US10003553B2 (en) | 2011-09-07 | 2018-06-19 | Konda Technologies Inc. | Optimization of multi-stage hierarchical networks for practical routing applications |
US10050904B2 (en) | 2009-10-16 | 2018-08-14 | Konda Technologies Inc. | VLSI layouts of fully connected generalized and pyramid networks with locality exploitation |
US20180285486A1 (en) * | 2017-03-29 | 2018-10-04 | International Business Machines Corporation | Method to synthesize a cross bar switch in a highly congested environment |
US10333508B2 (en) | 2017-03-29 | 2019-06-25 | International Business Machines Corporation | Cross bar switch structure for highly congested environments |
US10412025B2 (en) | 2013-07-15 | 2019-09-10 | Konda Technologies Inc. | Fast scheduling and optmization of multi-stage hierarchical networks |
US10536399B2 (en) | 2013-07-15 | 2020-01-14 | Konda Technologies Inc. | Automatic multi-stage fabric generation for FPGAs |
US11405331B1 (en) | 2011-09-07 | 2022-08-02 | Konda Technologies Inc. | Optimization of multi-stage hierarchical networks for practical routing applications |
US11405332B1 (en) | 2011-09-07 | 2022-08-02 | Konda Technologies Inc. | Fast scheduling and optimization of multi-stage hierarchical networks |
US11451490B1 (en) | 2009-10-16 | 2022-09-20 | Konda Technologies Inc. | VLSI layouts of fully connected generalized and pyramid networks with locality exploitation |
US11860814B1 (en) | 2020-11-01 | 2024-01-02 | Konda Technologies Inc. | Scalable distributed computing system with deterministic communication |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US3701112A (en) * | 1971-06-04 | 1972-10-24 | Bell Telephone Labor Inc | Balanced, incomplete, block designs for circuit links interconnecting switching network stages |
US4833468A (en) * | 1987-10-14 | 1989-05-23 | Unisys Corporation | Layered network |
US4988993A (en) * | 1986-12-23 | 1991-01-29 | At&T Bell Laboratories | Rearrangeable multiconnection switching networks constructed using combinatorial designs |
US5153843A (en) * | 1988-04-01 | 1992-10-06 | Loral Corporation | Layout of large multistage interconnection networks technical field |
US5471468A (en) * | 1992-06-01 | 1995-11-28 | Telefonaktiebolaget Lm Ericsson | Square switching architecture |
-
1997
- 1997-10-22 US US08/955,685 patent/US6018523A/en not_active Expired - Lifetime
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US3701112A (en) * | 1971-06-04 | 1972-10-24 | Bell Telephone Labor Inc | Balanced, incomplete, block designs for circuit links interconnecting switching network stages |
US4988993A (en) * | 1986-12-23 | 1991-01-29 | At&T Bell Laboratories | Rearrangeable multiconnection switching networks constructed using combinatorial designs |
US4833468A (en) * | 1987-10-14 | 1989-05-23 | Unisys Corporation | Layered network |
US5153843A (en) * | 1988-04-01 | 1992-10-06 | Loral Corporation | Layout of large multistage interconnection networks technical field |
US5471468A (en) * | 1992-06-01 | 1995-11-28 | Telefonaktiebolaget Lm Ericsson | Square switching architecture |
Cited By (46)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6185220B1 (en) * | 1998-06-15 | 2001-02-06 | Lucent Technologies, Inc. | Grid layouts of switching and sorting networks |
US7787449B2 (en) | 1999-06-08 | 2010-08-31 | Cisco Technology, Inc. | Butterfly network with switches set for two node disjoint paths and method for forming the paths |
US7113506B1 (en) * | 1999-06-08 | 2006-09-26 | Cisco Technology, Inc. | Butterfly network with switches set for two node disjoint paths and method for forming the paths |
US20070070993A1 (en) * | 1999-06-08 | 2007-03-29 | Feng Cao | Butterfly network with switches set for two node disjoint paths and method for forming the paths |
US6721311B1 (en) | 1999-08-10 | 2004-04-13 | Colorado Seminary | Self-routing permutation networks based on de Bruijn digraphs |
US20130124918A1 (en) * | 2002-12-02 | 2013-05-16 | Marvell World Trade Ltd. | Self-reparable semiconductor and method thereof |
US8812905B2 (en) * | 2002-12-02 | 2014-08-19 | Marvell World Trade Ltd. | Self-repairable semiconductor and method thereof |
US20040155725A1 (en) * | 2003-02-06 | 2004-08-12 | Com Dev Ltd. | Bi-planar microwave switches and switch matrices |
US6951941B2 (en) | 2003-02-06 | 2005-10-04 | Com Dev Ltd. | Bi-planar microwave switches and switch matrices |
US20070025257A1 (en) * | 2003-09-30 | 2007-02-01 | Fang Xu | Efficient switching architecture with reduced stub lengths |
US7863888B2 (en) * | 2003-09-30 | 2011-01-04 | Teradyne, Inc. | Efficient switching architecture with reduced stub lengths |
US20090100314A1 (en) * | 2007-10-15 | 2009-04-16 | Coreoptics Inc. | Modification of error statistics behind equalizer to improve inter-working with different fec codes |
US8555132B2 (en) * | 2007-10-15 | 2013-10-08 | Cisco Technology, Inc. | Modification of error statistics behind equalizer to improve inter-working with different FEC codes |
US8874988B2 (en) | 2007-10-15 | 2014-10-28 | Cisco Technology, Inc. | Modification of error statistics behind equalizer to improve inter-working with different FEC codes |
US20100180048A1 (en) * | 2009-01-09 | 2010-07-15 | Microsoft Corporation | Server-Centric High Performance Network Architecture for Modular Data Centers |
US9288134B2 (en) | 2009-01-09 | 2016-03-15 | Microsoft Technology Licensing, Llc | Server-centric high performance network architecture for modular data centers |
US8065433B2 (en) * | 2009-01-09 | 2011-11-22 | Microsoft Corporation | Hybrid butterfly cube architecture for modular data centers |
US9674082B2 (en) | 2009-01-09 | 2017-06-06 | Microsoft Technology Licensing, Llc | Server-centric high performance network architecture for modular data centers |
US10129140B2 (en) | 2009-01-09 | 2018-11-13 | Microsoft Technology Licensing, Llc | Server-centric high performance network architecture for modular data centers |
US20110080855A1 (en) * | 2009-10-01 | 2011-04-07 | Hei Tao Fung | Method for Building Scalable Ethernet Switch Network and Huge Ethernet Switch |
US8358597B2 (en) * | 2009-10-01 | 2013-01-22 | Hei Tao Fung | Method for building scalable Ethernet switch network and huge Ethernet switch |
US10554583B2 (en) | 2009-10-16 | 2020-02-04 | Konda Technologies Inc. | VLSI layouts of fully connected generalized and pyramid networks with locality exploitation |
US11811683B1 (en) | 2009-10-16 | 2023-11-07 | Konda Technologies Inc. | VLSI layouts of fully connected generalized and pyramid networks with locality exploitation |
US10050904B2 (en) | 2009-10-16 | 2018-08-14 | Konda Technologies Inc. | VLSI layouts of fully connected generalized and pyramid networks with locality exploitation |
US11451490B1 (en) | 2009-10-16 | 2022-09-20 | Konda Technologies Inc. | VLSI layouts of fully connected generalized and pyramid networks with locality exploitation |
US10977413B1 (en) | 2009-10-16 | 2021-04-13 | Konda Technologies Inc. | VLSI layouts of fully connected generalized and pyramid networks with locality exploitation |
US20110202682A1 (en) * | 2010-02-12 | 2011-08-18 | Microsoft Corporation | Network structure for data center unit interconnection |
US8665727B1 (en) | 2010-06-21 | 2014-03-04 | Xilinx, Inc. | Placement and routing for a multiplexer-based interconnection network |
US8415976B1 (en) | 2010-06-21 | 2013-04-09 | Xilinx, Inc. | Optimized interconnection networks |
US8098081B1 (en) | 2010-06-21 | 2012-01-17 | Xilinx, Inc. | Optimization of interconnection networks |
US7982497B1 (en) | 2010-06-21 | 2011-07-19 | Xilinx, Inc. | Multiplexer-based interconnection network |
US8493090B1 (en) | 2010-06-21 | 2013-07-23 | Xilinx, Inc. | Multiplexer-based interconnection network |
US10979366B1 (en) | 2011-09-07 | 2021-04-13 | Konda Technologies Inc. | Optimization of multi-stage hierarchical networks for practical routing applications |
US10992597B2 (en) | 2011-09-07 | 2021-04-27 | Konda Technologies Inc. | Fast scheduling and optimization of multi-stage hierarchical networks |
US10003553B2 (en) | 2011-09-07 | 2018-06-19 | Konda Technologies Inc. | Optimization of multi-stage hierarchical networks for practical routing applications |
US10574594B1 (en) | 2011-09-07 | 2020-02-25 | Konda Technologies Inc. | Optimization of multi-stage hierarchical networks for practical routing applications |
US10965618B1 (en) | 2011-09-07 | 2021-03-30 | Konda Technologies Inc. | Automatic multi-stage fabric generation for FPGAs |
US11777872B1 (en) | 2011-09-07 | 2023-10-03 | Konda Technologies Inc. | Optimization of multi-stage hierarchical networks for practical routing applications |
US11405332B1 (en) | 2011-09-07 | 2022-08-02 | Konda Technologies Inc. | Fast scheduling and optimization of multi-stage hierarchical networks |
US11405331B1 (en) | 2011-09-07 | 2022-08-02 | Konda Technologies Inc. | Optimization of multi-stage hierarchical networks for practical routing applications |
US10536399B2 (en) | 2013-07-15 | 2020-01-14 | Konda Technologies Inc. | Automatic multi-stage fabric generation for FPGAs |
US10412025B2 (en) | 2013-07-15 | 2019-09-10 | Konda Technologies Inc. | Fast scheduling and optmization of multi-stage hierarchical networks |
US10169511B2 (en) * | 2017-03-29 | 2019-01-01 | International Business Machines Corporation | Method to synthesize a cross bar switch in a highly congested environment |
US20180285486A1 (en) * | 2017-03-29 | 2018-10-04 | International Business Machines Corporation | Method to synthesize a cross bar switch in a highly congested environment |
US10333508B2 (en) | 2017-03-29 | 2019-06-25 | International Business Machines Corporation | Cross bar switch structure for highly congested environments |
US11860814B1 (en) | 2020-11-01 | 2024-01-02 | Konda Technologies Inc. | Scalable distributed computing system with deterministic communication |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US6018523A (en) | Switching networks having improved layouts | |
EP0463746B1 (en) | Segmentation of tracks in FPGA routing channels | |
KR100211123B1 (en) | Multi-stage interconnecting network for high speed packet switching | |
KR920004917B1 (en) | Self-routing switching circuit | |
US4899335A (en) | Self routing packet switching network architecture | |
US6473827B2 (en) | Distributed multi-fabric interconnect | |
KR950013157B1 (en) | Switching networks | |
JPH0766718A (en) | Wafer scale structure for programmable logic | |
US20030212877A1 (en) | Irregular network | |
US7468982B2 (en) | Method and apparatus for cluster interconnection using multi-port nodes and multiple routing fabrics | |
JPH0638256A (en) | Method for selection of connecting route of cross connection network | |
Jain et al. | TESH: A new hierarchical interconnection network for massively parallel computing | |
Koppelman et al. | A self-routing permutation network | |
US6091723A (en) | Sorting networks having improved layouts | |
US4910730A (en) | Batcher-banyan network | |
Lee et al. | A new decomposition algorithm for rearrangeable Clos interconnection networks | |
US6021131A (en) | Sorting networks having enhanced layout | |
Kawamoto et al. | The minimum width routing of a 2-row 2-layer polycell-layout | |
Hui | Switching integrated broadband services by sort-banyan networks | |
US7065076B1 (en) | Modular scalable switching networks | |
WO2000031609A1 (en) | A clock tree topology | |
US6327261B1 (en) | Translation process for an ATM cell header | |
US5216420A (en) | Matrix sorting network for sorting N inputs onto N outputs | |
US6088353A (en) | Sorting networks having reduced-area layouts | |
EP0186595B1 (en) | Routing technique |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: LUCENT TECHNOLOGIES INC., NEW JERSEY Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:EVEN, SHIMON;REEL/FRAME:009495/0929 Effective date: 19980924 |
|
FEPP | Fee payment procedure |
Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
AS | Assignment |
Owner name: THE CHASE MANHATTAN BANK, AS COLLATERAL AGENT, TEX Free format text: CONDITIONAL ASSIGNMENT OF AND SECURITY INTEREST IN PATENT RIGHTS;ASSIGNOR:LUCENT TECHNOLOGIES INC. (DE CORPORATION);REEL/FRAME:011722/0048 Effective date: 20010222 |
|
FEPP | Fee payment procedure |
Free format text: PAYER NUMBER DE-ASSIGNED (ORIGINAL EVENT CODE: RMPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
FPAY | Fee payment |
Year of fee payment: 4 |
|
AS | Assignment |
Owner name: LUCENT TECHNOLOGIES INC., NEW JERSEY Free format text: TERMINATION AND RELEASE OF SECURITY INTEREST IN PATENT RIGHTS;ASSIGNOR:JPMORGAN CHASE BANK, N.A. (FORMERLY KNOWN AS THE CHASE MANHATTAN BANK), AS ADMINISTRATIVE AGENT;REEL/FRAME:018590/0047 Effective date: 20061130 |
|
FPAY | Fee payment |
Year of fee payment: 8 |
|
FPAY | Fee payment |
Year of fee payment: 12 |
|
AS | Assignment |
Owner name: OMEGA CREDIT OPPORTUNITIES MASTER FUND, LP, NEW YORK Free format text: SECURITY INTEREST;ASSIGNOR:WSOU INVESTMENTS, LLC;REEL/FRAME:043966/0574 Effective date: 20170822 Owner name: OMEGA CREDIT OPPORTUNITIES MASTER FUND, LP, NEW YO Free format text: SECURITY INTEREST;ASSIGNOR:WSOU INVESTMENTS, LLC;REEL/FRAME:043966/0574 Effective date: 20170822 |
|
AS | Assignment |
Owner name: WSOU INVESTMENTS, LLC, CALIFORNIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:ALCATEL LUCENT;REEL/FRAME:044000/0053 Effective date: 20170722 |
|
AS | Assignment |
Owner name: WSOU INVESTMENTS, LLC, CALIFORNIA Free format text: RELEASE BY SECURED PARTY;ASSIGNOR:OCO OPPORTUNITIES MASTER FUND, L.P. (F/K/A OMEGA CREDIT OPPORTUNITIES MASTER FUND LP;REEL/FRAME:049246/0405 Effective date: 20190516 |
|
AS | Assignment |
Owner name: OT WSOU TERRIER HOLDINGS, LLC, CALIFORNIA Free format text: SECURITY INTEREST;ASSIGNOR:WSOU INVESTMENTS, LLC;REEL/FRAME:056990/0081 Effective date: 20210528 |